JP2006352896A - Wireless communication apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology of facilitating mounting by avoiding the processing from being complicated in the case of enhancing the throughput by aggregation of communication frames while maintaining quality of service of communication. <P>SOLUTION: The wireless communication apparatus includes: a main queue for storing MAC frames; a plurality of sub-queues related to the main queue and used for retransmission control by each priority; a means for extracting the MAC frame from the main queue on the basis of a destination and a priority identifier and distributing the extracted MAC frame to any of the sub-queues by each priority; and a means for extracting the MAC frame from the sub-queues and creating a MAC super-frame. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a wireless communication apparatus that performs medium access control, and more particularly, to access control for improving quality of service (QoS).

  Media access control (MAC) determines how a plurality of communication apparatuses that communicate by sharing the same medium use the medium to transmit communication data. In medium access control, as a result of two or more communication devices transmitting communication data using the same medium at the same time, the number of events (collisions) in which communication data on the receiving side cannot be separated is minimized. On the other hand, this is a technique for controlling access from a communication device to a medium so that the number of events in which the medium is not used by any communication device despite the presence of a communication device having a transmission request is minimized.

  Furthermore, some access control for improving quality of service (QoS) is also known. For example, there is HCCA (HCF Controlled Access), which is an extension of the conventional polling procedure, as QoS that guarantees parameters such as specified bandwidth and delay time. In HCCA, scheduling in consideration of required quality is performed in a polling procedure so that parameters such as bandwidth and delay time can be guaranteed.

Patent Document 1 refers to the QoS of the IEEE802.11e standard, and discloses a method for assigning priority to communication between wireless network stations.
JP 2002-314546 A

  Within the HC in IEEE802.11e, there is a scheduling processing unit that controls the transmission of polling frames to QSTA and the timing of transmitting downlink data. The scheduling processing unit transmits a polling frame and data for each priority so as to satisfy the required service quality using a traffic stream (TS) setup from QSTA.

  When a request from the scheduling processing unit in the HC makes it necessary to transmit data to a certain QSTA, the frame is stored in the retransmission sub-queue using the destination and priority as a key, and transmitted as a MAC super frame. A method is conceivable. Here, frames of multiple priorities for the same destination can be stored in one subqueue and transmitted as a MAC super frame, but sliding window control for each priority using a partial ACK response from the destination terminal The processing at the time of performing the process becomes complicated, and in some cases, the processing performance inside the apparatus may be reduced.

  The present invention has been made in view of such circumstances, and avoids complicated processing in improving throughput by aggregation of communication frames while maintaining communication quality of service (QoS). The purpose is to facilitate implementation.

  A wireless communication apparatus according to an aspect of the present invention includes a main queue for accumulating MAC frames, a plurality of subqueues associated with the main queue and used for retransmission control for each priority, a destination from the main queue, A radio comprising: means for extracting a MAC frame based on a priority identifier and distributing the MAC frame to one of the plurality of subqueues for each priority; and means for extracting a MAC frame from the plurality of subqueues and creating a MAC superframe. It is a communication device.

  According to the present invention, it is possible to provide a wireless communication apparatus capable of improving throughput by aggregation of communication frames while maintaining communication quality of service (QoS).

  Hereinafter, exemplary embodiments of the present invention will be described with reference to the drawings.

  A wireless communication apparatus according to an embodiment of the present invention is an apparatus that communicates with another communication apparatus via a wireless link, and includes processing units corresponding to a physical layer, a MAC layer, and a link layer. These processing units are realized as analog or digital electronic circuits according to mounting, or as firmware executed by a CPU incorporated in an LSI. An antenna is connected to the physical layer processing unit. The processing unit of the MAC layer has an aggregation processing unit. The aggregation processing unit includes a carrier sense control unit, a retransmission control unit, and a scheduling control unit.

First Embodiment [1-1. Frame Aggregation Implementation at HCCA (Sub Queue for Each Priority)]
When using IEEE802.11e HCCA (HCF controlled channel access) and performing MAC frame aggregation, if there is only one sub-queue for retransmission, the queue configuration is as shown in FIG. Here, CfQ is the data in the downlink (downlink) from the QoS access point (HC: Hybrid Coordinator) that can be transmitted during the polling period to the QoS terminal (QSTA: QoS Station), or the uplink from QSTA to the HC This is a main queue for accumulating (uplink) data, and there are various destination and priority MAC frames.

  As shown in FIG. 2, the HC in IEEE802.11e has a scheduling processing unit that controls transmission of a polling frame to QSTA and transmission of downlink data. It is assumed that the scheduling processing unit transmits a polling frame and data for each priority so as to satisfy the required service quality using a traffic stream (TS) setup from QSTA.

  When a request from the scheduling processing unit in the HC makes it necessary to transmit data to a certain QSTA, as shown in FIG. 1, the frame is stored in the retransmission sub-queue using the destination and priority as keys. Then, a method of transmitting as a MAC super frame is conceivable. Here, as shown in FIG. 3, it is possible to store frames of a plurality of priorities for the same destination in one subqueue and transmit them as MAC super frames, but using a partial ACK response from the destination terminal, The processing when performing the sliding window control for each priority is complicated, and in some cases, the processing performance inside the apparatus may be deteriorated.

  Therefore, in the present embodiment, as shown in FIG. 4, by preparing a plurality of retransmission control subqueues for each priority, sliding window control for each priority and frame aggregation for each priority can be realized more easily. It is also possible to perform parallel processing inside the communication device. First, when a request for downlink transmission with a certain priority (high priority in the example in the figure) occurs from the scheduling processor in the HC to a certain destination (STA1 in the example in the figure), the aggregation processing unit The destination and priority frames corresponding to the above are taken out from the main queue and stored in sub-queues prepared for each priority. The number of MPDUs (MAC Protocol Data Units) that can be aggregated in one MAC superframe is determined in advance by negotiation, and the upper limit is 8 in the example of FIG. Note that the negotiation method is not limited to a specific method.

  When downlink transmission of high priority frames to STA1 is performed as in the example of FIG. 4, the number of frames stored in the high priority subqueue is less than the maximum number that can be aggregated in the MAC super frame. , Store MAC frames of the same destination with different priorities in the corresponding sub-queues. In the example of FIG. 4, the maximum number of aggregates is 8, whereas only five high priority frames requested by the scheduling processing unit are stored in the subqueue. Store up to three frames in the medium priority subqueue. As shown in the figure, a sequence number is assigned for each priority according to the IEEE802.11e standard. Here, the request from the scheduling processing unit inside the HC is a method of “destination, priority” (extracting MAC frames from the main queue until the maximum number that can be aggregated to the MAC superframe is reached) or “destination , "Priority, number of frames" (the number of MAC frames specified by the scheduling processor is taken out from the main queue), "destination, priority 1, frame number of priority 1, priority 2, priority The method of “number of frames of 2” (transmission of a plurality of priorities and designation of the respective numbers from the scheduling processing unit) can be considered. Do the same.

  When actually creating a MAC super frame, the MAC frame is extracted with priority from the high priority sub-queue and is aggregated from the front of the MAC super frame. When considering the case of the example in FIG. 5, it is assumed that five MAC frames are stored in the high priority sub-queue and three MAC frames are stored in the medium priority. The total of these frames is equal to the maximum number that can be aggregated into one MAC super frame (8 in the example in the figure), but if there is no frame with the corresponding destination and priority in the main queue The maximum number that can be aggregated does not have to be reached. In FIG. 5, for example, when there are only two medium priority MAC frames in the main queue, a lower priority frame is stored in the low priority sub-queue and is placed behind the MAC super frame. Aggregate. The reason for the aggregation for each priority is to protect the high-priority MPDU in consideration of the fact that the longer the physical frame, the lower the channel estimation accuracy in the second half of the frame and the more likely it is to make errors. Yes.

  Note that this embodiment can be applied not only to downlink traffic from HC but also to uplink traffic transmission from QSTA.

Second Embodiment [1-2. Frame Aggregation Implementation at HCCA (Partial ACK Bitmap for Each Priority)]
For MPDUs aggregated for each priority in the MAC superframe, the high priority frame group is the forefront, the medium priority frame group is the second, the low priority frame group is the third, etc. As described above, a case is considered in which the relative order in which the priorities are aggregated is determined in advance, and this is recognized by both the transmitting and receiving terminals. As shown in FIG. 6, the MAC super frame receiving terminal writes the reception status of the aggregated MPDU in the partial ACK bitmap and returns a partial ACK. At this time, the priority type and arrangement order of the aggregated MPDUs do not affect the creation of the partial ACK bitmap. When a partial ACK is returned to the MAC super frame transmitting terminal side, bitmap information for each priority in the partial ACK bitmap is obtained from the number of MAC frames accumulated for retransmission in the subqueue for each priority. Can be judged. By extracting the transmission status for each priority, the sliding window control for each priority can be performed more efficiently.

  Here, as shown in FIG. 7, when the MPDU aggregated in the MAC super frame is not necessarily divided for each priority, even if a partial ACK is received from the destination terminal, the transmission status of each priority is changed. It is difficult to judge immediately. As a solution to this problem, there may be a method in which the MAC superframe transmission side has cache information indicating which priority is aggregated in which location. However, as shown in FIG. 8, the partial ACK bitmap is prioritized. A method of preparing each time is also conceivable. The expanded partial ACK frame has a “Number of Priority” field and a partial ACK bitmap for each priority. The “Number of Priority” field indicates the number of priorities existing in the partial ACK, and the “TID” field corresponds to a value of a priority identifier (TID: Traffic Identifier) of IEEE802.11e. As shown in the figure, in the extended partial ACK frame, there is a partial ACK bitmap for each TID. The number of “TID” fields and “partial ACK bitmaps” in the partial ACK frame is variable according to the number of priorities aggregated in the MAC super frame. In the example of FIG. This shows a case where there are three types of priority MPDUs in the superframe. In the example in the figure, the maximum number of MPDUs that can be aggregated in the MAC super frame is 8, and the field size of the partial ACK bitmap is 1 octet. This size also depends on the maximum number of aggregates. And may be variable.

  Here, as shown in FIG. 9, it is assumed that the MPDU aggregated in the MAC super frame is an error as a result of the CRC calculation. At this time, it becomes impossible for the MAC super frame receiving side to determine which priority is in what reception state. Therefore, as shown in FIG. 10, a “TID Bitmap” field is newly added to the MAC super frame header. The “Num of TID” field indicates the number of priorities aggregated in the MAC super frame. Following the “TID” field (the length is 1 octet, of which 4 bits are assigned to TID and the remaining 4 bits are assigned to reserved fields), a “TID Bitmap” field exists. The “TID Bitmap” field is information on how many frames of the priority are aggregated in the MAC super frame. By including this bitmap information in the MAC super frame header, when a partial error of MPDU other than the header CRC error occurs, which priority is present on the MAC super frame receiving terminal side, It is also possible to determine how the reception status was.

  FIG. 11 shows an example of using “TID Bitmap”. As shown in the figure, in the MAC super frame, "Medium priority" "High priority" "Medium priority" "High priority" "Low priority" "High priority" "High priority" "Low priority" Consider a case where MPDUs with three types of priority are aggregated. In the example of this figure, the maximum number of MPDUs that can be aggregated is set to 8, but of course this number is not fixed. At this time, the high priority “TID Bitmap” becomes “0 1 0 1 0 1 1 0”, the medium priority “TID Bitmap” becomes “1 0 1 0 0 0 0 0”, and the low priority “TID Bitmap”. “Bitmap” is indicated as “0 0 0 0 1 0 0 1”. That is, the identification information indicates where the MPDU corresponding to the priority exists in the MAC super frame.

  As the format of “TID Bitmap”, if the “Bitmap Information” field as shown in FIG. 12 is used, the “Number of TID” field can be omitted. The length of the “TID Bitmap” field in the MAC super frame header corresponds to the maximum number of MPDUs that can be aggregated, but the maximum number of aggregates is recognized by both the sending and receiving terminals through negotiation in advance. The receiving side can also determine the length of the “TID Bitmap” field. In the example of FIG. 13, three types of priority MPDUs are aggregated in the MAC super frame header with the maximum number of aggregates 8, and the length of the “TID” field in the MAC super frame header is a fixed length. In this case, the length of 1 octet and “TID Bitmap” field is also 1 octet (the maximum number of aggregates is 8), so the length of the “Bitmap Information” field (value in octet units described in the “Length” field) Is 6 octets. In the “Bitmap ID” field, an identifier for “TID Bitmap” is described, which is 2 in the example of FIG. 12, but it goes without saying that the value is not limited to this value.

  FIG. 14 shows a flow of communication by the MAC super frame in which the “TID Bitmap” field is expanded and the partial ACK in which the partial ACK bitmap for each priority is expanded in this embodiment. As shown in the figure, when MPDUs with multiple priorities are aggregated in the MAC super frame, the “TID Bitmap” field in the MAC super frame header describes the identification information of the location where the priority exists. Is done. In the example shown in the figure, “1 is present” is defined, but of course, it can be realized by negative logic. The MAC super frame receiving terminal performs the header CRC calculation of the MAC super frame header. If the header is not an error, the MAC super frame receiving terminal performs the CRC calculation of each aggregated MPDU. It is also easy to create a partial ACK bitmap for each priority level by using “TID Bitmap” in the MAC super frame header. As shown in the figure, regarding the creation of a partial ACK bitmap for each priority, it can be determined that there are four high-priority MPDUs from the “TID Bitmap”, and the two forward MPDUs are errors as a result of the CRC calculation. Therefore, the bitmap configuration is “0 0 1 1 0 0 0 0”. Since the length of the partial ACK bitmap is specified in units of octets, as a result, in the example of FIG. 14, the front 4 bits are information indicating the reception status.

  If a partial ACK that expands the partial ACK bitmap for each priority is received at the MAC superframe transmitting terminal side, procedures such as deletion of the MAC frame from the retransmission subqueue prepared for each priority are performed in parallel. Therefore, more efficient processing can be realized.

(Third embodiment) [1-3. Implementation of frame aggregation during HCCA (sliding window for each priority)]
When a partial ACK frame having a partial ACK bitmap for each priority is received, the MAC frame stored in the sub-queue prepared for each priority is deleted from the queue. As shown in FIG. 15, when MAC frames are stored in the high-priority and medium-priority subqueues, the MAC frame that can be correctly transmitted to the destination with reference to the partial ACK bitmap for each priority. Will be deleted. Here, if it is assumed that MPDUs in the MAC superframe are aggregated by dividing each priority, and the higher priority frame is packed in the front part, it is stored in the subqueue for each priority. The contents of the partial ACK bitmap are interpreted from the number of MAC frames that have been assigned, and the MAC frames of the respective priorities are deleted from the subqueue.

  The sliding window control is performed for each priority. Here, consider the cases as shown in FIGS. In these figures, two types of MPDUs of high priority and medium priority are aggregated in the MAC super frame. However, in implementing the present invention, the number of priorities is not limited. . In these figures, W_all defines the maximum number of MAC frames that can be transmitted continuously. The window W (high) (medium) at each time point indicates the maximum number of MPDUs that can be aggregated at one priority at a time, and the window slides backward depending on the situation of the partial ACK bitmap. Regardless of the priority, the example in the figure assumes that the maximum number that can be aggregated in the MAC superframe is 8. Further, FIGS. 16 and 17 show a case where there is physically one buffer on the receiving side, and FIGS. 18 and 19 show a sliding window when a plurality of buffers on the receiving side are physically prepared for each priority. It becomes control. Providing a plurality of physical buffers for each priority has an advantage that processing can be executed in parallel in the receiver. When there are a plurality of reception buffers for each priority, the MAC super frame transmission terminal determines each window W according to the reception buffer size for each priority of the destination terminal. The present invention is not limited to a specific negotiation method for window size determination.

  16 to 19, the notation “Null” means that the frame to be aggregated does not exist in the main queue, and the notation “Zero” correctly receives the previously sent frame at the destination. What can be done, and “NoAdd”, because of the maximum number of frames that can be aggregated in one MAC super frame, even if there is a corresponding frame in the main queue (if it does not exist), it will not be newly packed (do not pack) ) Means that.

  First, the sliding window control in the case where there is only one buffer on the MAC super frame receiving side will be described based on the examples of FIGS. The window size W (high) at each time point in FIGS. 16 and 17 is determined based on the physical buffer (single) of the destination terminal. The window size W (medium) has a variable length according to the number of aggregates of high priority frames.

  In FIG. 16, it is assumed that five high priority MPDUs and three medium priority MPDUs are aggregated and transmitted. Here, if five high priorities are aggregated, the window size W1 (medium) can be secured to a size equal to three because the remainder is less than the window size W1 (high). When these MPDUs having a plurality of priorities are aggregated, it is desirable to divide each priority and aggregate the higher priorities. Thereafter, it can be seen from the partial ACK information from the destination that all the high-priority MPDUs are transmitted normally, and that the medium-priority MPDU has the leading MPDU in error. In W2 (High), the high priority uses a sliding window, but the medium priority first frame (the first of the three sent) is incorrect, so the medium priority 2 in the destination terminal's buffer. Confirm that the frames are stored. Therefore, only five MPDUs, Seq6 to 10, can be newly aggregated with high priority. High priority cannot be packed any further, so “NoAdd” is displayed. The medium priority does not change the size of the window, and only the retransmission frames of Seq1 are subject to aggregation.

  Next, consider the example of FIG. The operation up to TX1 is the same as in FIG. It can be seen from the partial ACK from the destination that the second MPDU with the higher priority is incorrect, and that all the medium priority MPDUs have been transmitted normally. After performing sliding window control, if there are no MAC frames in the main queue that should be packed to high priority, the medium priority window size can be increased. For medium priority, we understand that 3 high priority MPDUs are stored in the buffer on the receiving side, so 4 frames are received together with 1 high priority frame for retransmission. You must think that it is filled with a side buffer (single). Therefore, the remainder obtained by subtracting 4 from 8 of the window size W2 (high) becomes the size of the window size W2 (medium). (The size of the window size (high) is a fixed length in this case) That is, it is possible to aggregate the four frames of Seq4 to 7 into the MAC superframe.

  The retransmission control as described above is performed, and MPDUs are aggregated within the range of the number indicated by W_all. In addition, aggregate frames are added with priority given to high-priority MPDUs.

  Consider FIG. 18 and FIG. 19 in the case where a plurality of physical buffers are provided for each priority on the receiver side. When a plurality of physical buffers are prepared, parallel processing can be performed inside the wireless communication apparatus. In this case, the buffer size on the receiving side is notified to the MAC super frame transmitting terminal for each priority, and the size of the window size W is determined for each priority on the transmitting side. In FIG. 18, it is assumed that five high-priority MPDUs and three medium-priority MPDUs are aggregated and transmitted in the MAC super frame as shown from TX1 (high) (medium). TX1 (high) Null means that there is no high-priority frame for a certain destination. In the example of FIG. 18, only five high-priority frames are stored in the main queue at the time of TX1. It has been shown. According to the contents of the partial ACK bitmap created at the time of RX1 (high) (medium), it is understood that retransmission is necessary for one medium priority frame after deleting the MAC frame from the subqueue. Thereafter, the window W (high) (medium) at each time point is slid for each priority. If there is only one priority aggregated to the MAC super frame, it is possible to pack the frames to the end of the window W at each time point, but there is one frame to be retransmitted with medium priority Therefore, in TX2 (high), seven frames from Seq. No. 6 to 12 are targets to be aggregated into the MAC super frame. Since TX2 (medium) has reached the maximum number of frames that can be aggregated into one MAC superframe, the total of seven high priority frames and one retransmitted at medium priority has reached the maximum. Do not add.

  As shown in FIG. 19, as a result of performing the sliding window processing after receiving the partial ACK, in TX2 (high), the high-priority frame aggregates only two frames for retransmission, so the medium-priority TX2 (Medium) can aggregate 6 frames from Seq. No. 4 to 9 (maximum amount of medium priority buffer specified on the receiving side). For each priority, the number of MPDUs to be continuously transmitted is determined so that the window size W falls within the range of W_all.

(Fourth Embodiment) [2-1. Block Ack Processing Efficiency]
Block ACK defined by IEEE802.11e supports efficient transmission by Selective Repeat. FIG. 20 shows a standard immediate block ACK sequence. As shown in the figure, transmission of block ACK in HCCA is adjusted within a channel use period (TXOP: Transmission Opportunity) designated by a QoS access point (HC: Hybrid Coordinator). The example of FIG. 20 shows a state where polling to QSTA or data transmission to the downlink is performed within a CAP (Controlled Access Period) period. The HC in FIG. 20 transmits a QoS CF-Poll frame (a QoS-compliant polling frame transmitted by the HC to permit transmission to the QSTA) to the QSTA1. QSTA1 can freely transmit a frame as long as it is within the range of a given TXOP. In FIG. 20, the QoS data targeted for block ACK is transmitted in bursts to QSTA2, and the period ends. Thereafter, during the TXOP period 2, the HC transmits QoS data targeted for block ACK in bursts at SIFS intervals to QSTA2. QSTA1 in TXOP period 3 transmits a block ACK request to QSTA2 and waits for block ACK. Further, in TXOP period 4, HC transmits a block ACK request to QSTA2 and waits for block ACK.

  Here, with IEEE802.11e block ACK, in order to create a block ACK bitmap indicating the reception status on the receiver side, up to 1024 reception histories for 64 MSDU (MSDU: MAC Size Data Unit) are stored. It is necessary to manage for each destination and TID (Traffic Identifier). In the current specification, there is no provision that a block ACK request must be transmitted immediately after QoS data transmitted in bursts from a certain destination. Therefore, every time a block ACK request is received, the destination (and TID) is received. Checking status and creating block ACK generally increases the processing load on the receiver side. As a result, there may be a case where only the transmission by the delay type block ACK as shown in FIG. 21 can be realized in time for the SIFS period. The delayed block ACK is a block ACK that is transmitted after a certain time has elapsed after receiving the block ACK request, and it is clear that the transmission efficiency decreases. This is because when block ACK transmission executed within the transmission period of one terminal straddles the next TXOP period, QoS data of block ACK transmission from another terminal is interrupted, and each reception status is managed on the receiving side This is due to the complicated mechanism.

  Therefore, in the first configuration example of the present embodiment, the terminal that has obtained the TXOP includes all the series of sequences from QoS data burst transmission, block ACK request transmission, and block ACK reception within that period. Perform scheduling. FIG. 22 shows a frame sequence. In the block ACK sequence of FIG. 20, even if a certain TXOP period is given, the block ACK request is not necessarily transmitted within the TXOP period. However, in the present embodiment shown in FIG. Schedule the burst transmission, transmission of block ACK request, and reception of block ACK without fail in the terminal. Specifically, the block ACK request is transmitted with a margin so that the number of QoS data to be transmitted in bursts is reduced and the block ACK can be received reliably. As for the timing of block ACK request transmission, an appropriate timing is calculated from the duration, number, and physical transmission rate of each QoS data. Note that the present invention does not limit the calculation method of the block ACK request transmission timing to a specific method. In addition, within a certain TXOP period, QoS data of multiple priorities (TID) can be transmitted in bursts, but in this case as well, measures such as limiting the number of QoS data to be burst transmitted, As shown in FIG. 23, frame transmission scheduling is performed so that a block ACK for each TID can be received in TXOP. Furthermore, as shown in FIG. 24, transmission by block ACK may be performed for a plurality of destinations within the TXOP period, but transmission is also performed so that block ACK from each destination can be received within the TXOP period. Schedule frame transmission on the terminal side. If there is only time left to transmit one QoS data and block ACK request cannot be transmitted, data transmission assuming block ACK to the destination (or TID) at the next opportunity (TXOP ) Is preferred.

  Next, in the second configuration example of the present embodiment, a block ACK request is aggregated and transmitted to the last QoS data that is burst-transmitted in a given TXOP period. As shown in FIG. 25, the channel data can be used more effectively by aggregating and transmitting the QoS data and the block ACK request in one physical frame. 26 and 27 show frame formats of aggregated QoS data and a block ACK request. In the example of FIG. 26, the aggregate is based on information such as “Type”, “Sub Type”, “Length”, etc. in the MAC header of IEEE802.11 (which may include information for QoS control of IEEE802.11e). 3 shows a format in which information indicating the delimiter (length) of the MPDU is provided as MAC payload. This field is called an aggregation field. Also, an FCS (Frame Check Sequence) for a field indicating the length of the 802.11 MAC header and MPDU is added. Bitmap information or the like for the purpose of function expansion can be added to the aggregation field. FIG. 27 shows an example in which an identification header indicating the length of the aggregated MPDU is newly provided. This header is called an aggregation header. A header CRC for calculating an error in the header is added to the aggregation header. If the header is incorrect, all the aggregated frames are discarded. This aggregation header is added in front of the data payload such as aggregated QoS data and block ACK request. Note that the aggregation header format in FIG. 27 is the case where the maximum number that can be aggregated is 8. Therefore, it is of course possible to aggregate a plurality of QoS data as shown in FIG. It is also possible to aggregate only QoS data subject to block ACK instead of a request. Although the maximum number of MPDUs that can be aggregated in one physical frame needs to be recognized in advance between the transmitting and receiving terminals through some negotiation, a specific negotiation method is out of the scope of the present invention.

  Here, as shown in FIG. 29, only QoS data can be aggregated into one physical frame and transmitted in bursts. As shown in FIG. 30, even when QoS data having a plurality of priorities is transmitted in bursts within a certain TXOP period, transmission is performed by aggregating and transmitting a block ACK request for each of a plurality of TIDs to the last QoS data. Efficiency can be improved. Further, if the block ACK for each TID is aggregated and transmitted from the destination terminal, the effect is greater. Here, as shown in FIG. 31, a plurality of QoS data burst-transmitted at SIFS intervals in FIG. 30 can be aggregated and transmitted as one physical frame.

  Next, the receiver side of QoS data by block ACK will be described.

  FIG. 32 is a block diagram of a wireless communication apparatus corresponding to the receiver side of QoS data by block ACK. The receiver has a control unit that receives QoS data to be transmitted in bursts, a block ACK request, and a control unit that creates a block ACK. Furthermore, it has a storage area for immediate block ACK and a storage area for delayed block ACK.

  When the bursty reception of the QoS data is started from a certain destination, the reception status of the data frame is stored in the immediate block ACK storage area. This area secures 1024 reception statuses for each TID. 1024 is a value obtained by multiplying the maximum number of MSDUs (MAC Size Data Units) that can be continuously transmitted within the block ACK period × the maximum number of fragment frames per MSDU. As shown in FIG. 22, if a block ACK request is received from the destination storing the reception status in the immediate block ACK storage area by the end of the TXOP period, a block ACK response is immediately returned. Since QoS data from multiple destinations is not mixed, it is possible to reduce the processing load of creating a block ACK for one destination.

  Further, as in the example of FIG. 20, while QSTA2 is continuously receiving burst ACK target QoS data from HC in a burst manner, the next TXOP period starts and blocks from other destinations QSTA1 When reception of ACK QoS data is started, if there is not enough room in the immediate block ACK storage area of FIG. 32, the reception status information for HC stored therein is represented as a delayed block ACK. The reception status information of the new destination (QSTA1) is created in the immediate block ACK storage area. Thereafter, a response with a delayed block ACK is made to the old destination (HC in the figure).

  As described above, according to the present embodiment, it is possible to improve the processing efficiency of Block Ack, and to reduce the processing load on the receiver side regarding the creation of Block ACK.

  In EDCA, RTS-CTS is used to inform other terminals of the channel usage period used by itself, but the duration (line usage period) in the RTS frame includes QoS data, block ACK request, and block ACK. The TXOP calculation is performed so that the series of sequences up to this point is contained.

(Fifth Embodiment) [3-1. Frame Aggregation in EDCA]
In EDCA (Enhanced Distributed Channel Access), a plurality of access categories (AC) are provided for each priority, and they perform carrier sense back-off control in parallel. Each AC has its own IFS (AIFS) period, and the higher the priority AC, the shorter the carrier sense of the time that can be started. If transmission processing between multiple ACs occurs simultaneously within the MAC, a high-priority AC frame is sent to the PHY, and the low-priority AC increases the CW (contention window) and then repeats random Enter back-off waiting time. The state of channel access in EDCA is shown in FIG. Further, as expressed in FIG. 34, in the EDCA, data from an upper layer is mapped to TID (Traffic Identifier) Nos. 0 to 7 in accordance with User Priority of IEEE802.1D. The mapped TID is distributed to AC. As a result, a plurality of TID MAC frames are stored in the AC.

  In EDCA, when a plurality of MPDUs are aggregated in one physical frame, they are usually packed into one destination and one TID as shown in FIG. This is because EDCA is a contention-based access control method (Prioritized QoS) and does not guarantee the quality of each traffic stream like HCCA. That is, a MAC frame is extracted from the head of the main queue prepared for each AC, and frames having the same destination, TID are aggregated. In this case, as can be seen from FIG. 35, it may be less than the maximum number of MPDUs that can be aggregated in one physical frame, and the channel utilization efficiency may not be fully utilized.

  Therefore, in this embodiment, when a certain AC acquires a transmission right and transmits a data frame, the MAC frame is extracted from the head of the main queue for each AC, and the maximum number that can be aggregated into one physical frame is not reached. In this case, the MAC frame of the other TID (EDCA stipulates that there are two types of TID stored in the AC) is to be aggregated. At this time, it is desirable to prepare for each TID a sub-queue for storing MAC frames taken out from the main queue for each AC for retransmission. By preparing a sub-queue for each TID, the sliding window can be controlled more easily.

  As can be seen from FIG. 34, there is a difference in priority depending on the TID even within the same AC. Therefore, in this embodiment, for example, as shown in FIG. Channel estimation in OFDM (estimating phase and amplitude distortion in the transmission path for each subcarrier) is realized using a known preamble signal stored in the receiver. In a wireless LAN that is communication in the packet mode and in which the time variation of the transmission path in the packet (frame) is small, a method of performing channel estimation at the head of the preamble signal independently for each packet is common. However, when the frame length increases as in the case of an aggregated frame, the transmission path fluctuates in time, so that the estimation result calculated at the time of preamble reception may not be accurately reflected toward the latter half of the frame (FIG. 37). ). Therefore, by aggregating high priority MPDUs in the forward part, an effect of increasing error tolerance of TID data frames with high priority can be obtained.

  In addition, since sequence numbers are assigned consecutively for each TID, the wireless communication apparatus that transmits the aggregated frame performs sliding window control for each TID by a partial response from the receiving side.

  Until now, multiple TID frames within the same AC have been aggregated, but there is no other TID MAC frame within a certain AC, and as a result of the collision avoidance mechanism, internal When the transmission timings overlap, MAC frames with the same destination may be taken out from the AC main queue having a lower priority than the AC and aggregated into one physical frame. In this case, since an internal collision occurs in the low priority AC, it is desirable to increase the contention window and take backoff after transmission of the aggregated frame.

  As described above, according to the present embodiment, it is possible to enhance the effect of frame aggregation during EDCA.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

Diagram showing conventional queue configuration Block diagram showing HC device configuration (MAC layer) Diagram showing subqueues and multiple priority frames The figure which shows the subqueue for every priority which concerns on embodiment of this invention Diagram for explaining creation of MAC superframe The figure which shows the case where the order of priority is recognized beforehand Diagram showing the case where priority order is indefinite Diagram showing an extended partial ACK frame The figure for demonstrating judgment of CRC error and priority Diagram showing the addition of “TID Bitmap” Diagram showing usage example of “TID Bitmap” Diagram showing “Bitmap Information” format Diagram showing the use of Bitmap Information The figure for demonstrating judgment of CRC error and priority Diagram showing frame deletion from sub-queues by priority The figure which shows an example of the sliding window control for every some priority which concerns on embodiment of this invention The figure which shows the other example of sliding window control for every multiple priorities The figure which shows an example of sliding window control when a receiving side has a buffer for every priority The figure which shows the other example of sliding window control in case a receiving side has a buffer for every priority Diagram showing an example of block ACK sequence (immediate type) Diagram showing delayed block ACK sequence The figure for demonstrating the block ACK scheduling in TXOP which concerns on embodiment of this invention Diagram showing scheduling of multiple priorities in TXOP Diagram showing scheduling of multiple destinations in TXOP The figure which shows the 1st aggregate example of the last QoS data and a block ACK request | requirement The figure which shows the 1st aggregate example of QoS data and a block ACK request | requirement The figure which shows the 2nd aggregate example of QoS data and a block ACK request | requirement The figure which shows the 3rd aggregate example of QoS data and a block ACK request | requirement The figure which shows the 2nd aggregate example of the last QoS data and a block ACK request | requirement The figure which shows the 3rd aggregate example of the last QoS data and the block ACK request | requirement of multiple priorities A diagram for explaining the efficiency of block ACK by aggregation The figure which shows the reception status storage area for immediate type block ACK and delayed type block ACK The figure which shows the mode of channel access in EDCA Diagram showing mapping between AC and User Priority The figure for explaining the problem at the time of frame aggregation at the time of EDCA The figure which shows the frame aggregation of multiple TID at the time of EDCA which concerns on embodiment of this invention The figure which shows the channel estimation precision along a time-axis in the aggregation for every high priority TID

Explanation of symbols

HC ... QoS access point (Hybrid Coordinator), QSTA ... QoS terminal (QoS Station)

Claims (18)

In a wireless communication device that performs transmission by block ACK through a permitted channel usage period (TXOP: Transmission Opportunity),
Transmission means for performing burst transmission of QoS data and transmission of a block ACK request;
A wireless communication apparatus comprising: adjustment means for adjusting burst transmission of the QoS data and transmission of a block ACK request so that a block ACK can be received within the channel use period.   The radio communication apparatus according to claim 1, wherein the adjustment unit calculates a value of TXOP so that the block ACK can be received within the TXOP and notifies the other terminal of the value.   The radio communication apparatus according to claim 1 or 2, wherein the last QoS data frame and the block ACK request frame transmitted at SIFS intervals within a certain channel use period are aggregated into one physical frame and transmitted.   The wireless communication apparatus according to claim 1, wherein when a plurality of MPDUs are aggregated in one physical frame, information indicating the length of each MPDU is included in the MAC frame payload. In a wireless communication apparatus that performs channel access using EDCA (Enhanced Distributed Channel Access),
Means for creating a single physical frame in which MAC frames of a plurality of TIDs (Traffic Identifiers) existing in each AC (Access Category) are aggregated;
A wireless communication apparatus comprising: a transmission unit configured to transmit the physical frame to a destination terminal.   The wireless communication apparatus according to claim 5, wherein the MAC frames are aggregated by dividing the MAC frames in a descending order of TID priority.   The wireless communication apparatus according to claim 5 or 6, wherein a partial response indicating a reception status of the physical frame at a destination terminal of the physical frame is received, and sliding window control is performed for each TID according to the partial response. A main queue for accumulating MAC frames;
A plurality of sub-queues associated with the main queue and used for retransmission control for each priority;
Means for taking out a MAC frame from the main queue based on the destination and priority identifier, and distributing the MAC frame to one of the plurality of sub-queues for each priority;
Means for extracting a MAC frame from the plurality of sub-queues and creating a MAC super frame. An access point scheduler that performs HCCA (HCF controlled channel access);
The wireless communication apparatus according to claim 8, further comprising: a unit that transmits the MAC super frame in response to a request from the scheduler.   9. The wireless communication apparatus according to claim 8, further comprising means for transmitting the MAC super frame in response to polling from an access point performing HCCA (HCF controlled channel access).   The wireless communication apparatus according to claim 8, wherein a copy of a MAC frame is extracted from the plurality of subqueues, aggregated into one MAC super frame, and transmitted.   The wireless communication apparatus according to claim 8, wherein a copy of a MAC frame is extracted from the plurality of subqueues, aggregated and transmitted in one physical frame for each priority.   11. The MAC super frame header includes aggregating MAC frames having a plurality of priorities and including identification information indicating how many of the priorities are packed in the MAC super frame header. Wireless communication device.   The wireless communication apparatus according to claim 8, wherein the wireless communication apparatus receives a partial ACK frame having bitmap information indicating a reception status for a plurality of priority MAC frames aggregated in the MAC super frame for each priority.   11. When receiving a partial ACK frame including a partial ACK bitmap for each priority, the MAC frame is deleted from the sub-queue for each priority by using the partial ACK bitmap for each priority. The wireless communication device described.   The wireless communication apparatus according to claim 8, wherein when a plurality of buffers are physically provided on the MAC super frame receiving terminal side, the window size is managed and aggregated for each priority.   When partial retransmission of a MAC super frame occurs and sliding window control is performed for each priority, the number of MAC frames to be retransmitted among all priorities, and a MAC newly added with a priority higher than that priority The number of MAC frames to be newly aggregated is determined for each priority in consideration of the number of frames, and the retransmission MAC frame and the newly added MAC frame are aggregated into one physical frame and transmitted. The wireless communication device described. Means for receiving a MAC superframe in which a plurality of priority MAC frames are aggregated;
Means for transmitting a partial ACK frame having bitmap information indicating the reception status for the plurality of priority MAC frames for each priority.

Images